J. Comp. Neurol. 3374:555-577.   



Direct Catecholaminergic-Cholinergic Interactions in the Basal Forebrain. II. Substantia Nigra-Ventral Tegmental Area Projections to Basal Forebrain Cholinergic Neurons





Center for Molecular and Behavioral Neurosciences, Rutgers University, Newark NJ 07102 (R.P.G., L.Z.), and Department of Pharmacology, Vrije Universiteit, Amsterdam, The Netherlands (R.P.G.)



Running headline: Dopaminergic-Cholinergic Interactions


Key words: PHA-L, double immunolabeling, correlated light and electron microscopy, 3-D reconstruction


41 pages, 11 figures


Associate Editor: P. E. Sawchenko


*Author for correspondence:

Laszlo Zaborszky, M.D., Ph.D.

Center for Molecular and

Behavioral Neurosciences

Rutgers University,

197 University Avenue

Newark, NJ 07102, U.S.A.

TEL: 201-648-1080/Ext. 3181

Fax 201-648-1272


            Previous observations indicate that the basal forebrain receives dopaminergic input from the ventral midbrain. The present study aimed at determining the topographic organization of these projections in the rat, and if this input directly terminates on cholinergic neurons. Injections of the anterograde tracer Phaseolus vulgaris-leucoagglutinin (PHA-L) into discrete parts of the ventral tegmental area (VTA) and the substantia nigra pars compacta (SNC) labeled axons and terminals in distinct parts of the basal forebrain, including medial and lateral septum, diagonal band nuclei, ventral pallidum, globus pallidus, substantia innominata, globus pallidus, and internal capsule, where PHA-L-labeled terminals abutted cholinergic (choline acetyltransferase=ChAT-containing) profiles. 3-D computerized reconstruction of immunostained sections clearly revealed distinct, albeit overlapping, subpopulations of ChAT-immunoreactive neurons apposed by PHA-L-labeled input from medial VTA (mainly in vertical and horizontal diagonal band nuclei), lateral VTA and medial SNC (ventral pallidum and anterior half of substantia innominata), and lateral SNC (caudal half of the substantia innominata and globus pallidus). At the ultrastructural level, about 40% of the selected PHA-L-labeled presynaptic terminals in the ventral pallidum and substantia innominata were found to establish synaptic specializations with ChAT-containing profiles, most of which on the cell body and proximal dendritic shafts. Convergent synaptic input of unlabeled terminals that formed asymmetric synapses with the ChAT-immunoreactive profiles were often found in close proximity to the PHA-L-labeled terminals. These observations show that the cholinergic neurons in the basal forebrain are targets of presumably dopaminergic SNC/VTA neurons, and suggest a direct modulatory role of dopamine in acetylcholine release in the cerebral cortical mantle.




            The mammalian basal forebrain contains large cholinergic neurons which give rise to an extensive and continuous network of cholinergic innervation of the cerebral cortex, hippocampus, olfactory bulb, and amygdala. These projection neurons are distributed across several territories of the basal forebrain, extending from the medial septal pole rostrally, through the diagonal band nuclei and the ventral pallidum (VP), to the substantia innominata (SI) and the globus pallidus (GP) caudally. As demonstrated in a three-dimensional reconstruction of serial sections of the rat brain, the cholinergic neurons make up a continuous column (Schwaber et al., 1987) and are therefore considered to be a single functional unit, referred to as the basal forebrain cholinergic (BFC) system. 

            The BFC system has been the focus of considerable interest, since it has been implicated in aging-associated neurodegenerative disorders, including Alzheimer's and Parkinson's disease (Bowen et al., 1976; Davies and Maloney, 1976; Whitehouse et al., 1981; Perry et al., 1987). Although its precise function is unclear, the BFC is thought to play a modulatory role in complex behavioral functions, including arousal, attention, sensory processing, motivation, learning and memory (for review, see Dekker et al., 1991; Fibiger, 1991). Since the BFC neurons are widely distributed throughout the basal forebrain, it is likely that a certain level of differentiation with regard to input and output relationship may exist, dependent upon their location. Individual cholinergic projection neurons are likely to be related to the complex circuitries of the territories in which they are embedded. This may also imply differences in their precise role in the animal's behavior (for ref. see Zaborszky et al., 1991). A vast body of information on the topographic organization of the efferent projections of the BFC has emerged in recent years. The BFC can be roughly parcelled into subdivisions, each with its own characteristic set of major target regions, as shown with retrograde and anterograde tracing studies in rats (e.g.,Mesulam et al., 1983b; McKinney et al, 1983; Saper, 1984; Zaborszky et al., 1986a; Luiten et al., 1987, Gaykema et al., 1990) and monkeys (e.g., Mesulam et al., 1983a).

            Several tract-tracing studies have revealed a number of transmitter-specific systems in the brainstem which innervate basal forebrain areas (Beckstead et al., 1979; Fallon and Moore, 1978; Semba et al, 1988; Vertes, 1988; Jones and Cuello, 1989). Among these are the serotonergic median and dorsal raphe nuclei, noradrenergic locus coeruleus, and the dopaminergic substantia nigra pars compacta (SNC) and ventral tegmental area (VTA). These findings are substantiated by the presence of tyrosine hydroxylase (TH)- and dopamine-b-hydroxylase-immunoreactive fibers in the substantia innominata (Martinez-Murillo et al., 1988; Chang, 1989), ventral pallidum (Zaborszky, 1989) and adjacent areas (Freedman and Cassell, 1994;  Zaborszky and Cullinan, 1996). Neurochemical lesion of the SNC results in profound reduction of dopamine levels in the SI (Geula and Slevin, 1989). Furthermore, dopamine appears to be a functional neuromodulator in the pallidal areas and the SI (Napier et al., 1991). The basal forebrain may therefore be a site of monoaminergic-cholinergic interaction with functional significance, as has recently been shown neuropharmacologically for the dopaminergic influence on cholinergic activity (Day and Fibiger, 1992, 1993; Imperato et al., 1993). Disruption of dopaminergic/cholinergic interaction due to dysfunction in both dopaminergic and cholinergic systems may account for the cognitive disturbances in patients with Alzheimer's or Parkinson's disease (see Decker and McGaugh, 1991).

            When searching for a better understanding of the fundamental mechanisms of transneuronal interactions between the BFC neurons and their dopaminergic afferents, the question emerges as to whether dopaminergic fibers terminate directly on these cholinergic cells or relay their input exclusively through a complex circuitry of local interneurons. The topographic relationship between DAergic fibers and the BFC neurons has not previously been addressed directly, nor has the contribution from various parts of the SN-VTA. Choline acetyltransferase (ChAT)-TH double-labeling immunocytochemistry has revealed direct associations between TH-positive terminals and cholinergic neurons in the globus pallidus (see joint paper of Zaborszky and Cullinan). Since TH is not a conclusive marker for dopaminergic neurons, it was questioned if the ventral midbrain dopaminergic nuclei indeed terminate in the basal forebrain cholinergic cell areas. Therefore, we sought to determine the distribution of efferents from the SN and VTA (A9-A10 DAergic cell groups of Dahlström and Fuxe, 1964) to the basal forebrain, and to see if these axons establish synaptic contacts with cholinergic neurons. Ascending projections to the basal forebrain were demonstrated with the anterograde tracer Phaseolus vulgaris-leucoagglutinin (PHA-L). Furthermore, we investigated direct SN-VTA input on BFC neurons using correlated light- and electron microscopic analysis.



Animal surgery


            Twenty-five male Sprague-Dawley rats, weighing 250-300 grams, were used in this study. Prior to surgery, the rats were anesthetized with sodium pentobarbital (50 mg/kg body weight). A 2.5% PHA-L (Vector Labs, Burlingame, CA) solution in 0.1 M sodium phosphate buffer, (pH 8.0) was backfilled into glass micropipettes with tip diameter of 10-20 µm and stereotaxically placed into the SN-VTA in the right hemisphere. The stereotaxic coordinates were chosen according to the atlas of Paxinos and Watson (1986). PHA-L was iontophoretically delivered for 15-20 minutes at 6 µA with a 7 seconds on/7 seconds off pulse (Gerfen and Sawchenko, 1984). In nine cases with SN injections, the electrode was held at two sites along the vertical injection track spaced 0.3 mm apart, for 15 minutes per injection. This was done in order to improve the chance of tracer deposition in the thin, compact dopamine cell layer of the SNC.


Tissue preparation


            After the appropriate survival period (5-10 days), the animals were given an overdose of pentobarbital. They were then briefly perfused transcardially with normal saline (0.9% NaCl, 1 minute), followed immediately by 500 ml of cold fixative. The first 350 ml of the fixative was composed of 4% paraformaldehyde (PF), 0.1% glutaraldehyde (GA) and 15% saturated picric acid in 0.1 M phosphate buffer (PB, pH 7.4). GA was omitted in the subsequent 150 ml of fixative. After the perfusion, the brains were removed and further fixed overnight in the same fixative at 4˚C.

            After fixation, transverse sections (50 µm thick) were Vibratome-cut from the forebrain (septal pole to caudal globus pallidus) and the mid- and hindbrain (including the SN-VTA), and were collected into six series in ice-cold 0.1 M PB. Subsequently, the sections selected for electron microscopy were immersed in 15% and then 30% sucrose, exposed to two freeze-thaw cycles within liquid nitrogen to improve antibody penetration, and extensively washed in PB.




            Prior to all incubation steps, the sections were rinsed several times in ice-cold PB. Normal sera and antibodies were diluted in a PB solution to which 0.5% Triton X-100 had been added when processed for light microscopy, or in PB containing only 0.04% Triton X-100 when processed for electron microscopy.

            Localization of PHA-L. The first series of forebrain and brainstem sections was processed for the tracer PHA-L only, using the avidin-biotin-peroxidase (ABC) method. Sections were incubated overnight at 4˚C in goat anti-PHA(E+L) (Vector Labs, Burlingame, CA) at a dilution of 1:1000. This was followed by incubation in biotinylated donkey anti-goat IgG (Jackson Immunoresearch Labs) at 1:100 for 2 hours, and the ABC complex (Vector Labs) at 1:500 for 2 hours. Subsequently sections were treated with the coupled oxidation reaction of Itoh et al. (1979) with 3,3'-diaminobenzidine (DAB) as a chromogen intensified with ammonium nickel sulfate (black reaction product). The solution was made up in PB and contained 50 mg DAB, 40 mg ammonium chloride, 0.4 mg glucose oxidase (Sigma, type VII), 200 mg ß-D-glucose/100 ml, and 1 mM ammonium nickel sulfate. The reaction medium was filtered before use. The sections were rinsed three times after the staining, mounted on slides, dehydrated and coverslipped in DPX. The brainstem sections were counterstained with Cresyl violet to determine the exact location of the PHA-L injection sites.

            Simultaneous localization of PHA-L and TH. Another series of brainstem sections were processed for immunofluorescence detection of PHA-L and tyrosine hydroxylase (TH), to relate the distribution of the PHA-L-labeled neurons to the dopaminergic cells of the SN-VTA, and to assess the proportion of double-labeled cells. Every third section through the SN-VTA was incubated in goat anti-PHA(E+L) (1:1000, overnight, in PB with 0.5% Triton), followed by biotinylated donkey anti-goat IgG (1:100, 2 hours) and Texas Red-Streptavidin (Jackson, 1:200, 2 hours). The sections were then incubated in rat anti-TH (1:200, overnight), and in Fluorescein-conjugated goat anti-rat IgG (Jackson, 1:100, 2 hours). Sections were mounted, coverslipped and viewed with a Zeiss Axioplan epifluorescent microscope with appropriate filter set.

            Simultaneous localization of PHA-L and ChAT. The second series of forebrain sections was processed to visualize PHA-L and ChAT with a sequential double-labeling method. The first part (PHA-L immunolabeling) was similar to that described for the first series of sections (above). PHA-L was visualized with nickel-intensified DAB. The sections were then treated with the second immunocytochemical protocol. This involved incubation in a rat monoclonal anti-ChAT antibody (Eckenstein and Thoenen, 1982) at 1:10 for 2 overnights at 4˚C, followed by donkey anti-rat IgG (Jackson Immunoresearch Labs) at 1:100 for 2 hours, rat monoclonal peroxidase-anti-peroxidase (PAP) (Sternberger-Meyer) at 1:100 for 2 hours, and finally development using DAB alone as a substrate, again by the coupled glucose oxidation reaction. Sections were rinsed, dehydrated, and coverslipped.

            The possibility of crossreactivity between immunoreagents in double-labeling experiments was controlled by omitting of one of the primary antibodies, or by replacing with normal serum. Spare series of sections were run for this purpose. Evidence of cross reactivity, however, was never encountered.


Electron microscopy


            The third series of forebrain sections was used for electron microscopy and was processed for PHA-L and ChAT double immunolabeling. The sequence and the antibody solutions were the same as described previously, except that biotinylated anti-PHA(E+L) (Vector Labs, dilution 1:200) was used and that the incubation times of the secondary antibodies, ABC and PAP, were 3-4 hours. The entire procedure was carried out at low temperature (4˚C).  Following this immunostaining procedure, the sections were postfixed in 1% osmium tetroxide for 30-45 minutes, dehydrated in an ascending ethanol series (1% uranyl acetate was included in the 70% ethanol step for 40 minutes) and propylene oxide, and flat-embedded in  Durcupan  ACM  (Fluka)  between  glass  slides  and  coverslips  coated with liquid releasing agent (EMS). The embedded sections were examined under the light microscope (Zeiss Axioplan). Selected areas containing PHA-L-labeled varicosities apparently contacting ChAT-immunolabeled cell bodies or dendrites in the basal forebrain were photographed, dissected and mounted with cyanoacrylate on the flat surface of cylindrical resin blocks, trimmed and ultrathin-sectioned on a Reichert Ultracut E ultramicrotome with a Diatome diamond knife. Serial ultrathin sections were collected on single-slot Formvar-coated gilded grids and examined with a Philips 201 electron microscope. Micrographs were taken at 60 keV.


Analysis of the material


            The PHA-L injections sites were evaluated by plotting PHA-L filled cell bodies from coronal midbrain sections at magnifications between 4x and 20x with the aid of the Neurolucida image analysis set-up (MicroBrightField Inc., Colchester, VT) connected to a Zeiss Axioplan microscope.  Aiming for a detailed evaluation of the organization of the SN-VTA input in relation to the basal forebrain cholinergic projection system, we analyzed the PHA-L-ChAT double-labeled series of sections in two different ways. First, drawings of 3 representative cases were made at 10x with the aid of a camera lucida drawing tube at various basal forebrain levels. Both PHA-L-labeled fibers and ChAT-positive cell bodies and proximal dendritic segments were drawn. Second, sections from these cases were used to map at high resolution the putative contacts between PHA-L-labeled terminals and labeled cholinergic elements with the aid of the Neurolucida image analysis set-up. The outline and gross brain territories were drawn with 4x and 10x lenses. The cholinergic cell bodies were then mapped with the 40x lens. Finally, the basal forebrain area was systematically scanned using a 63x objective lens with oil immersion, and putative contact sites and recipient cell bodies were marked. The same criteria for selecting cases for analysis at the electron microscopic level, were used, i.e., a clearly identified PHA-L-labeled terminal swelling (including associated axon) directly abutting a ChAT-labeled profile, with both structures appearing in the same focal plane. As such, every sixth section was analyzed and their images were aligned and merged together yielding a three-dimension-like image of the forebrain, which could be viewed from different angles.



General pattern of PHA-L labeling


            The PHA-L injections were placed in different parts of the SN and VTA. Labeled neurons were confined to the VTA (n=7), the SN (n=13), or distributed in both (n=5). Two of the VTA injections were placed in the very medial cell column, covering the nucleus interfascicularis, the n. linearis rostralis, and n. linearis caudalis (see Fig. 1, case 93128). The other VTA injections were situated more laterally, mainly centered ventromedial to the medial lemniscus in the n. paranigralis. In five cases, both the ventromedial portion of the SNC and the adjacent lateral portion of the VTA contained labeled neurons, packed around the medial lemniscus (Fig. 1, case 93047, Fig. 2A). The injections in the SN proper were further subdivided based on their mediolateral distribution. Seven cases, however, resulted in limited fiber labeling in the striatum and virtually none in the basal forebrain. These were all characterized by PHA-L-labeled cell bodies confined to the SN pars reticulata, an area with few DAergic cells innervating the striatum (Gerfen et al., 1987). These cases were omitted from the analysis, but serve as evidence that all basal forebrain afferents found after SN injections arise from the SN pars compacta (SNC). The other cases showed abundant PHA-L labeling in neurons of the SN pars compacta, confined either to the medial (not shown), the middle (Fig. 1, case 93049, Fig. 2B), or the lateral portion of the SN (Fig. 1, case 93051).

            Only those cases with a large portion of PHA-L-labeled cells in DAergic cell-rich parts of the ventral midbrain resulted in massive anterograde labeling of fiber plexuses in portions of the dorsal and/or ventral striatum, as well as moderately dense axonal and terminal staining in basal forebrain regions rich in cholinergic neurons. Within the VTA and SNC, a large portion (between 30 and 70%) of the PHA-L-labeled cells also contained TH, as confirmed with PHA-L/TH double immunofluorescence labeling (Fig. 2C-F). Rough comparison between cases revealed a very poor correlation of topographic differences in projections with variations in rostrocaudal locations of the injections. In contrast, the location of the injection sites along the mediolateral axis appeared to be a far more important factor defining the pattern of striatal and basal forebrain labeling. The general patterns of basal forebrain projections from particular medial to lateral portions of the VTA and SN strongly resembled the distribution of labeling after injections of 3H-leucine/proline, as described by Fallon and Moore (1978) and Beckstead et al. (1979). There is also a dorsoventral differentiation in output of the DAergic neurons, i.e. the distinction between dorsal and ventral tiers (Gerfen et al., 1987). However, we could not draw any conclusion with regard to this mode of organization, since none of the injections in the SN was confined to a single tier. An exception was noted for the injections confined to the SN pars reticulata (SNR), which resulted in patchy labeling of the dorsal striatum as reported by Gerfen et al. (1987), but extremely sparse labeling in the basal forebrain. This could be interpreted as indicating that most of the basal forebrain innervation from the SN-VTA arises from the DA-containing neurons of the dorsal tier. Based on these observations, we have selected 4 cases with very distinct but characteristic patterns of anterograde labeling after injections in medial to lateral parts of the SN-VTA: case 93128 (injection in medial VTA), case 93047 (medial SN and adjacent lateral VTA), case 93049 (middle third of the SN), and case 93051 (lateral third of the SN). Figure 1 shows the actual distribution of labeled cells in each of these cases.


Distribution of ChAT-immunoreactive neurons


            The distribution of ChAT-immunopositive neurons in the sections processed for PHA-L-ChAT double-labeling, as indicated in red in figures 4, 5, and 6, was in full agreement with previous reports on localization of ChAT immunoreactivity (Eckenstein et al,., 1988; Wainer et al., 1984). Red-brown immunostained large neuronal cell bodies with long, sparsely branching dendritic extensions were distributed throughout the medial septum (MS), vertical and horizontal limbs of the diagonal band (VDB and HDB), anterior and subcommissural ventral pallidum (VP), subcommissural and sublenticular substantia innominata (SI), internal capsule, ventromedial and caudal globus pallidus (GP), and neighboring parts of the anterior amygdala and bed nucleus of the stria terminalis. Slightly smaller ChAT-positive cells were also evenly distributed throughout the caudate putamen, n. accumbens, and the olfactory tubercle. The latter group of neurons were not included in the analysis, as described below. ChAT-immunoreactive cells in the anterior part of the VP, which perforates through ventral striatal cell regions, could not easily be distinguished from their striatal counterparts as to whether they are projection neurons or belong to the population of striatal interneurons. These were not included in the analysis.




Patterns of SN-VTA afferents in relation to ChAT-containing neurons


            Fiber trajectories. Darkfield-illuminated micrographs of anterogradely labeled fibers and terminals (Fig. 3) show examples of innervation patterns in the ventral pallidum (VP), horizontal and vertical limb of the diagonal band nuclei (HDB, VDB). A detailed illustration of different patterns of PHA-L-labeled projections related to the distribution of cholinergic neurons is given in figures 4, 5, and 6. In general, PHA-L-labeled axons ascend in a rostral direction in the medial forebrain bundle (mfb) and internal capsule (ic). From the medial VTA, the fibers run through the medial component of the mfb to its most rostral extension, from which they innervate the rostromedial portions of the SI (Fig. 4A), the VP, HDB, bed nucleus of the stria terminalis (Fig. 4B), VDB, medial and lateral septal areas, the olfactory tubercle, and the nucleus accumbens (Fig. 4C,D). Fibers from the lateral VTA and medial SN run in rostral direction through a slightly more lateral component of the mfb, as well through the most medial corner of the ic. From the mfb, a contingent of fibers reach the HDB, VDB, and VP. A smaller portion ascends further to the septum.  The bundle in the medial ic spreads into a large portion of the sublenticular SI and ventromedial globus pallidus (GP) (Fig. 5A,B). In addition, fascicles emanate from the ic into the rostral half of the GP and reach the dorsomedial and ventromedial parts of the caudate putamen, the core of the nucleus accumbens, and the olfactory tubercle, which are densely innervated regions (partly omitted in Fig. 5B-D, indicated with asterisks). Following PHA-L injection in the middle third of the SN, labeled ascending axons are distributed in the far lateral part of the mfb and in a larger portion, also more lateral, of the ic (Fig. 6A,B). From there, they radiate into the GP and sublenticular SI at a more caudal level. Subsequently, more caudolateral portions of the SI, GP, and caudate putamen are innervated (Fig. 6B-D). From the lateral SN (e.g., case 93051, not shown), ascending fibers run in dense fascicles through even more caudal and lateral portions of the mfb and ic. These reach the caudalmost parts of the SI, GP, and the caudolateral part of the caudate putamen. When the injection involved the SNR, an additional bundle of axons was seen to run from the mfb/ic dorsomedially into the thalamus, forming a dense plexus in the ventromedial nucleus (Fig. 6A,B). Three representative cases, shown in figures 4, 5, and 6, will be described in more detail: 93128 (with injection in medial VTA, see Figs. 1, 2), 93047 (lateral VTA and medial SN), and 93049 (middle portion SN).

            Case 93128. Figure 4 shows the characteristic pattern of fibers and terminations from the medial VTA, together with the perikarya of the ChAT-containing neurons. Since PHA-L was deposited bilaterally in the medial VTA, anterograde labeling was similar in both hemispheres (the right hemisphere is shown in Fig. 4). Caudal levels of the BFC neurons are devoid of labeling, and therefore only the anterior half of the BFC system is charted. The labeled fibers which ascend through the mfb give off varicose branches in the lateral hypothalamic/substantia innominata transition area, and appose some of the medially situated ChAT-positive cells (Fig. 4A). Few fibers traverse in more lateral directions. At more rostral levels, profusely branching fibers form an increasingly dense network, covering the ChAT-positive-rich cell parts of the HDB and the subcommisural VP/SI (Fig. 4B). Numerous varicosities and boutons were encountered, some of which were apposed to ChAT-immunoreactive profiles. Farther rostral, axons reach the rostromedial portion of the VP (Fig. 4C). A particularly dense plexus is prominent, covering the dorsal ridge of the angular part of the VDB, where it merges into the HDB. This plexus can be followed rostrally to the most anterior part of the VDB (Fig. 4D). The most lateral group of cholinergic cells in the angular wing of the VDB was embedded in this terminal network.  Another dense plexus covers the dorsomedial part of the shell of the accumbens (Fig. 4D). A smaller contingent of fibers runs through the VDB and the medial septum. In these cholinergic cell-rich nuclei, relatively few varicose branches were seen (Fig. 4C,D). Other fibers run along the lateral edge of the VDB and medial septum to terminate in the medial half of the lateral septum.

            Case 93047. From cells in the lateral VTA and neighboring medial SN, labeled axons travel through the mfb and medial ic reaching the cholinergic cell-rich areas of the basal forebrain from caudal portions of the SI (not shown) to the rostral pole of the VDB (Fig. 5). A large number of fibers radiate from the mfb into virtually the entire extension of the SI, HDB, and adjacent regions, intermingling with ChAT-containing cells (Fig. 5A,B). A small number of fascicles emanate from the ic at caudal levels and traverse the caudal half of the GP, where they give off a few branches, which appose some of the caudal BFC cells there. They finally reach the medial rim of the caudate putamen bordering the GP (Fig. 5B). More rostrally, the fibers tend to bifurcate more frequently, revealing a diffuse plexus of varicose fibers in the sublenticular SI, HDB, and subcommisural VP, often apposing ChAT-positive profiles. The dense contingent of fibers in the ventromedial ic closely approaches the cholinergic cells, and a small number of apparent appositions of branches was found.  The largest part of the contingent of labeled fibers passes the ic and the GP and reaches the rostromedial portions of the striatum. Similar to case 93128 described above, there was a discrete but dense plexus of terminations within a small area at the medial border of the VP at the level of the crossing of the anterior commissure (Fig. 5B). In the rostral direction, a contingent of axons runs through the dorsal aspects of the HDB into the VDB up to its rostral pole (Fig. 5C,D). Profusely branching fibers bearing varicosities were found in close proximity to cholinergic cells of the rostromedial VP and dorsolateral aspects of the VDB, although these were not as dense as in the subcommisural VP and SI. The septum received only a few fibers.

            Case 93049. Following an injection in the middle, with some spread to the lateral part of the SN, a fiber distribution in the basal forebrain emerges which is essentially restricted to the sublenticular SI, the ic and the caudal portion of the GP (Fig. 6). Notably the ic is invested with labeled fascicles heading in the rostrolateral direction to central and lateral parts of the caudate putamen (Fig. 6A,B). Nevertheless, bifurcating fibers carrying varicosities ramify from the major trajectories into the SI, the ventromedial corner and the caudal third of the GP, and even within the ic (Fig. 6C-D). Here, appositions with cholinergic cells were encountered, although not as frequently as seen in the SI in case 93047. In rostral direction, the number of axons gradually thins out, leaving a limited number of varicose fibers in the subcommissural VP (Fig. 6D).

            Morphology of PHA-L labeled afferents. Smooth fibers of various thicknesses were abundant, found traveling through most of the basal forebrain sectors. Very thin afferents (type A according to Gerfen et al., 1987) were numerous in the striatal target regions, but were only occasionally encountered in the basal forebrain regions. Here, a slightly thicker type of afferent was most common, showing a more wrinkled appearance with varicosities usually between 0.5 and 1.0 µm in size and variable intervaricose segments (up to 15 µm). In addition, fibers were present with more closely packed varicosities, and drumstick-like boutons of a larger size (up to 2 µm) appeared on short branches.


Topography of PHA-L-ChAT appositions: 3-D reconstruction


            Using high resolution light microscopy, we screened the entire basal forebrain of the above-described cases  for close appositions between PHA-L-labeled terminals and ChAT-containing profiles. Figure 7 are 3-dimensional composite images of stacked sections of cases 93128, 93047, and 93049 for comparison. Most of the appositions were found on proximal dendritic elements, the cell body of which was present in the same section. These putative recipient cells are highlighted in red. The far majority of identified cells were apposed by one labeled terminal within the section. Some double or triple contacts were seen on the same cholinergic profile. For better visualization, in Figure 7, in the left column (A,C,E) the composite images are viewed from dorsal, while in the right column, the sections are viewed from lateral (60˚ from frontal).

             The images clearly reveal distinct, albeit overlapping, subpopulations of putative BFC recipient cells from the medial VTA (case 930128), those from the lateral VTA and medial SN (case 93047), and those from the middle and lateral thirds of the SN (case 93047). Numerical comparison of the PHA-L-ChAT contact sites is summarized in the diagram of Fig. 8. Input from medial to lateral portions of the VTA-SN was found in progressively more caudal (and lateral) parts of the BFC system. Each case shows an increase and subsequent decrease in the number of putative recipient ChAT-positive neurons along the rostrocaudal axis, with a maximum that is progressively more caudal along the rostrocaudal axis when PHA-L was injected more laterally in the VTA-SN. The maximal percentages varied  between 10 and 14% of all ChAT-immunoreactive cells present at this rostrocaudal level. A considerable overlap is apparent as well. The cholinergic cells apposed by the medial VTA (case 93128) and the lateral VTA-medial SN (case 93047) are larger in number and are distributed over a wider rostrocaudal range of the BFC than are those innervated by more lateral aspects of the SN (cases 93049 and 93051).

            The distribution of putative contact sites matches that of the fiber distribution as described above in each case. There may not be a simple relationship, however, between fiber density and incidence of appositions with cholinergic profiles. For instance, in case 93047, there is a strong tendency for afferents to appose cholinergic elements in the subcommisural VP, but not in the more ventral HDB, despite a similar density of profusely branching fibers in these regions (as shown in Fig. 5B).


Electron microscopic observations


            PHA-L-ChAT double labeling. A total of 34 PHA-L-labeled boutons abutting ChAT-immunopositive elements were selected for ultrastructural analysis on the basis of the same criteria used for the mapping as described above. Labeled terminals were encountered on cell bodies (n=5), proximal dendrites within 50 µm distance from the parent cell body (n=11), and distal dendritic shafts (n=18). As can be viewed in Figures 9, 10, and 11, the DAB reaction product in the cholinergic elements caused a flocculent or patchy electron dense labeling of the cytoplasm and adhered to the outer surfaces of membranes of subcellular organelles. In contrast, Nickel-enhanced DAB reaction product in the PHA-L-labeled terminals showed a much more homogeneous, and intensely electron-opaque precipitate filling the entire profile. Mitochondria and oval-shaped or round vesicles could be distinguished as being relatively electron-lucent. Figure 11D is a color micrograph of a plastic embedded section, showing that the two different immunochromogen labels (DAB versus Ni-DAB) could be distinguished even after embedding at the light microscopic level (reddish brown vs. black). Ten out of the 34 selected terminals appeared to make no direct contact with ChAT-positive neurons, either with proximal (n=3) or distal dendrites (n=7), but were separated by two or more layers of membranes of an interposed profile. The remaining 24 boutons were found indeed to abut directly against the apposed ChAT-immunolabeled structure. In some cases, however, locally poor ultrastructural preservation of the membranes or an oblique sectioning angle with the plane of the contact surface hampered the drawing of definitive conclusions about the presence of synaptic specializations. Synaptic membrane specializations could be identified with greater certainty on 14 contacts, reconstructed from 3 animals. Of these, 3 were on cell bodies (Fig. 9D, 10D, 11F), 6 on proximal (Figs. 11D,E) and 5 on distal dendritic shafts.       

            The synaptic specialization displayed in Figure 9D seem to be symmetric, in most of the cases, however, the  identification of the type of synapse was not possible because of the dense reaction product in the pre- and/or postsynaptic elements (Fig. 10F, 11H). In one case (Fig. 11G), the synapse was asymmetric, with a pronounced  postsynaptic density. A consistent feature noted was that the labeled presynaptic element forms part of a configuration in which the postsynaptic cholinergic element is simultaneously contacted by another, unlabeled axon, the latter often establishing an asymmetric synaptic junction. Occasionally, the two synaptic contacts were found next to each other (Fig. 9D,D').





            The results of this study demonstrate that the SN-VTA, presumably dopaminergic afferents terminate on basal forebrain cholinergic neurons, and also provide a detailed analysis of the topographic organization of different compartments of the SN/VTA-BFC connectivity. This unique and extensively combined analysis at both high resolution light microscopic and ultrastructural levels strongly suggests that, based on this input from the SN and VTA, the BFC system is a functionally heterogeneous system under strong differential dopaminergic influence.






Methodological considerations


            In this study, we used DAB and Ni-enhanced DAB as a combination of chromogens for localization of ChAT-positive structures and PHA-L-labeled projections. Both in high resolution light- and electron microscopic analysis, the two chromogens could readily be distinguished. Immunolabeling with NiDAB consistently resulted in a homogeneous distribution of extremely electron-dense precipitate, whereas the DAB precipitate showed a flocculent precipitate, permitting the identification of subcellular organelles. Moreover, the use of correlated light (color difference) and electron microscopy removes any doubt as to the antigenic origin of the different immmunolabels at the EM level. ChAT-immunoreactive terminals are very sparse in the areas studied (Ingham et al., 1985; Martinez-Murillo et al., 1990). In our study, ChAT terminals were occasionally encountered in the vicinity of the ChAT-positive postsynaptic profiles, but did not terminate on them. These terminals showed a much lighter, flocculent immunoprecipitate (not shown). Therefore, it is highly unlikely that the strongly immunopositive terminals were mistaken for projections from the VTA and SN. 

            With the selection criteria used, in about 40% of the cases we could identify with great certainty the synapse between the PHA-L-labeled varicosity and the cholinergic profile. The majority of these synapses were with cholinergic cell bodies and proximal dendrites (64%) as opposed to distal dendrites (36%), confirming the result of the accompanying paper (Zaborszky and Cullinan, 1996) that TH-positive appositions were most frequently observed at proximal portions of the cholinergic neurons. Assuming that PHA-L axon varicosities originating from adjacent midbrain dopaminergic cells terminate on different, but slightly overlapping population of cholinergic neurons, and furthermore, that about half of putative contact sites mapped at the light microscopy, may indeed represent true synaptic sites (above), it is safe to suggest that the proportion of dopaminoceptive cholinergic neurons could make up 5-20% of the total BFC system, depending on the rostro-caudal location (see Fig. 8).


Monosynaptic SN/VTA-BFC connectivity


            The present study using correlated light-electron microscopic analysis shows that there is a direct SN-VTA input on BFC neurons. There is, however, no conclusive evidence that the identified terminals on the cholinergic neurons are dopaminergic afferents, although this is very likely for two reasons. First, retrograde tracing of basal forebrain afferents combined with TH immunolocalization revealed that the great majority of neurons in the SN-VTA that give rise to basal forebrain projections are dopaminergic (Semba et al., 1988; Jones and Cuello, 1989; Martinez-Murillo et al., 1988). In addition, TH-immunoreactive and DA-containing fibers have been localized in virtually all components of the BFC system, i.e., the SI, GP, VP, diagonal band nuclei, and medial septum (Voorn et al., 1986; Martinez-Murillo et al., 1988; Jones and Cuello, 1989; Freedman and Cassell, 1994). Second, PHA-L-labeled terminals were found to contact BFC neurons in a similar fashion to the TH-positive appositions described in the accompanying report by Zaborszky and Cullinan. For example, in certain aspect of the globus pallidus no DBH-containing (i.e., noradrenergic) terminals occur and thus TH labeling likely represents dopaminergic axons, labeled TH-axons were in close association with cholinergic cell bodies and proximal dendrites. In contrast, DBH-positive terminals contacted mainly distal cholinergic dendrites (see accompanying paper of Zaborszky and Cullinan).  Nevertheless, the contribution of non-dopaminergic afferents to the projection on cholinergic neurons cannot be ruled out.

            Recently, microdialysis experiments in rats revealed that cortical acetylcholine release is indeed under dopaminergic control (Day and Fibiger, 1992, 1993). These studies showed that frontal cortical acetylcholine release was facilitated by DA through a D1 receptor mechanism, and this was paralleled by behavioral arousal of the rats. The site of interaction was not at the cholinergic terminal field in the cortex, but likely within the basal forebrain, although the exact localization is as yet unknown. An indirect pathway either through local GABA interneurons or through the nucleus accumbens (Zaborszky and Cullinan, 1992), which receives massive dopaminergic input, cannot be ruled out (Wood, 1986, Zaborszky et al., 1986b).  An indirect mechanism to control cholinergic activity cannot be ruled out, since the great majority of PHA-L labeled terminals in the basal forebrain were located adjacent to non-cholinergic profiles.  Indeed, using a similar correlated light- electron microscopic approach, we have found axosomatic contacts from the SN-VTA on parvalbumin-positive neurons in the VP and SI (Gaykema and Zaborszky, in preparation). Nevertheless, the present results provide an anatomical substrate for a direct dopaminergic/cholinergic interaction. More conclusive evidence of the site of dopaminergic control in the basal forebrain awaits colocalization studies of DA receptor subtypes (Levey et al., 1993) in this region with ChAT and other neurotransmitter-specific markers.


Topographic organization of the SN-VTA afferents


            Another conclusion of this study is that the SN-VTA afferents to the BFC are topographically organized, such that medial parts of the ventral midbrain innervate rostromedial components of the BFC system, whereas more lateral components project to more caudolateral parts. This topographical organization corresponds to the distribution of the major mesostriatal pathways in gradually more lateral portions of the medial forebrain bundle and internal capsule, when comparing medial vs. lateral subdivisions of the A9-A10 dopaminergic system (Beckstead et al., 1979; Fallon and Moore, 1978). Lateral divisions of the mesostriatal pathway radiate to  caudal terminal fields, whereas the most medial portion traverses the medial forebrain bundle to reach the rostral pole of the basal forebrain.

            Superimposed on this principle is a stronger and more widespread innervation of the BFC system from the VTA and the ventromedial portion of the SNC than from more lateral parts of the SN. As a consequence, there is an overlap of cholinergic cells receiving input from different parts of the ventral midbrain. However, our reconstruction technique cannot discriminate whether the same cholinergic neuron receives input from various portions of the SN-VTA or whether individual cholinergic cells receiving input from different portion of the ventral mesencephalon are intermingled. Nevertheless, the medial VTA mainly targets the diagonal band nuclei in the rostroventral portion of the basal forebrain, whereas lateral VTA and medial SNC project to more dorsocaudally located cholinergic cells, predominantly in the subcommissural VP and the anterior half of the SI. The middle and lateral parts of the SNC terminate on cells of the caudal third of the BFC system, distributed in the SI and GP. Since dopaminergic neurons of the SN-VTA are functionally segregated (Deutch et  al., 1993), this differential distribution of afferents from the different parts of the SN-VTA suggest that the various subdivisions of the basal forebrain cholinergic systems receives distinct dopaminergic input.

            It is noteworthy that the medial septum is not an important target of the SN-VTA, in contrast to the suggestion of a previous report (Milner, 1991). We were unable to encounter an appreciable number of close appositions between PHA-L-labeled fibers and cholinergic neurons in this region. This may be due, however, to the fact that we missed a particular area in the VTA or SN, which might provide a stronger input to the medial septum.

            The question arises as to whether the terminations in the cholinergic cell areas of the basal forebrain are collaterals of mesostriatal and/or mesocortical fibers, or originate in a specific subset of DAergic cells in the VTA and SNC.  Considering the fact that mesocortical and mesostriatal projections in rat arise from distinct neuronal populations in the VTA with restricted overlap (Deutch et al., 1993), and that Smith et al. (1989) have described a separate mesopallidal pathway in the squirrel monkey, it is likely that separate VTA dopaminergic cells project to the nucleus accumbens, the cholinergic forebrain projection neurons and the prefrontal cortex. However, at present no direct data is available. With respect to the biochemical heterogeneity of the DAergic VTA and SNC, the contribution to BFC input of cholecystokinin- and neurotensin-costoring DAergic neurons has yet to be determined.


Functional significance of SN/VTA-BFC connectivity


            The SN-VTA dopaminergic system is implicated in a wide array of behavioral aspects, including locomotion, motivation, mood, positive reinforcement and reward. In humans, this system is thought to play a major role in various affective disorders such as depression, schizophrenia, and drug abuse (see Fibiger and Phillips, 1986 for review). A major part of the DAergic mechansims in behavior is mediated through the projections to the nucleus accumbens, the dorsal striatum, and to the prefrontal cortex. Since the SN/VTA-BFC pathway forms a seemingly minor component in quantitative terms (as compared to the massive striatal projections), we can only speculate as to the specific functions the dopaminergic relay on cholinergic corticopetal pathways may subserve.

            It is interesting to note, that dopaminoceptive cholinergic neurons appear to be located in more extensive basal forebrain areas than those cholinergic neurons that receive noradrenergic input.  If one compares the location of catecholaminoceptive cholinergic neurons with data about  their presumptive target areas (for ref. see Zaborszky et al., 1991), it can be concluded that cholinergic cells projecting to allocortical areas seem to receive both dopaminergic and noradrenergic input; on the other hand, BFC neurons innervating neocortical areas come primarily under a dopaminergic influence.  In view of the fact that basal forebrain cholinergic cells receive topographical organized projections from the nucleus accumbens (Zaborszky and Cullinan, 1992) and  ventral mesencephalic dopaminergic neurons (this study), it is conceivable that basal forebrain cholinergic projection neurons might be part of the highly specific and parallel organized cortico-striato-pallido-thalamo-cortical and cortico-striato-pallido-mesencephalic-cortical loops (Deutch et al., 1993; Groenewegen and Berendsee, l994). Ultimately, an understanding of the impact of dopaminergic-cholinergic interactions in the basal forebrain in normal and diseases states might require interpretation in this context.

            The activity of the BFC neurons correlates with behavioral arousal and electro-encephalic desynchronization (Detari and Vanderwolf, l987; Buzsaki et al., 1988). Increased cortical activity also correlates well with increased acetylcholine release (Celesia and Jasper, 1966; Metherate et al., 1992), as does behavioral activation (Day et al., 1991), but causal relationship among behavior, EEG and BFC activity remains elusive. Recently, Day and Fibiger (1992, 1993), and Imperato et al. (1993) have shown that dopaminergic stimulation has a profound effect on cortical and hippocampal acetylcholine release, suggesting a critical role for DAergic afferents in mediating increased BFC tone in behavioral states of attention and arousal.

            In Alzheimer's disease and Parkinson's disease, both cholinergic and dopaminergic systems appear to be highly prone to degeneration. This has led to the idea of a possible causative relationship between the reduction of dopaminergic and cholinergic markers. Experimental lesioning of the DA-containing afferents to the basal forebrain has resulted in decreased levels of ChAT in the BFC system (Zaborszky et al., 1993; Robertson and Staines, 1994). Whether ChAT levels in BFC neurons are indeed dependent on DAergic input is still unclear, but further studies may reveal a new form of dependency of cholinergic neurons on their afferents for their structural, biochemical and functional integrity, in addition to the trophic role for the BFC of the target-derived trophic substances.



            This work was supported by USPHS grants NS23945 and NS30024. Dr. Lennart Heimer (University of Virginia) is gratefully acknowledged for providing the necessary electron microscope facilities. We greatly appreciate the excellent technical skills of Debra Swanson in electron microscopy.




Beckstead, R.M., V.B. Domesick, and W.J.H. Nauta (1979) Efferent connections of the substantia nigra and ventral tegmental area in the rat. Brain Res. 175: 191-217.


Bowen, D.M., C.B. Smith, P. White, and A.N. Davison (1976) Neurotransmitter‑related enzymes and indices of hypoxia in senile dementia and other abiotrophies. Brain 99: 459‑496.


Buzsáki, G., R.G. Bickford, G. Ponomareff, L.J. Thal, R. Mandel, and F.H. Gage (1988) Nucleus basalis and thalamic control of neocortical activity in the freely moving rat. J. Neurosci. 8: 4007‑4026.


Celesia, G.G., and H.H. Jasper (1966) Acetylcholine released from cerebral cortex in relation to state of activation. Neurology 16: 1053-1063.


Chang, H.T. (1989) Noradrenergic innervation of the substantia innominata: a light and electron microscopic analysis of dopamine-B-hydroxylase immunoreactive elements in the rat. Exptl. Neurol. 104: 101-112


Dahlstrom, A., and K. Fuxe (1964) Evidence for the existence of monoamine-containing neurons in the central nervous system. I. Demonstration of monoamines in the cell bodies of brain stem neurons. Acta physiol. Scand. 62:1-55.


Davies, P., and A.J.F. Maloney (1976) Selective loss of central cholinergic neurons in Alzheimer's disease. Lancet 2: 1403.


Day, J., and H.C. Fibiger (1992) Dopaminergic regulation of cortical acetylcholine release. Synapse 12: 281-286.


Day, J., and H.C. Fibiger (1993) Dopaminergic regulation of cortical acetylcholine release: effects of dopamine receptor agonists. Neuroscience 54: 643-648.


Day, J., G. Damsma, and H.C. Fibiger (1991) Cholinergic activity in the rat hippocampus, cortex and striatum correlates with locomotor activity: an in vivo microdialysis study. Pharmac. Biochem. Behav. 38: 723-729.


Decker, M.W., and J.L. McGaugh (1991) The role of inter­actions between the cholinergic system and other neuromodula­tory systems in learning and memory. Synapse 7: 151-168.


Dekker, A.J.A.M., D.J. Connor, and L.J. Thal (1991) The role of cholinergic projections from the nucleus basalis in memory. Neurosci. Biobehav. Rev. 15: 299-317.


Detari, L., and C.H. Vanderwollf (1987) Activity of identified cortically projecting and other basal forebrain neurones during large slow waves and cortical activation in anesthetized rats. Brain Res. 437: 1-8.


Deutch, A.Y., A.J. Bourdelais, and D.S. Zahm (1993) The nucleus accumbens core and shell:  accumbal compartments and their functional attributes. In P.B. Kalivas and Ch. D. Barnes (eds): Limbic Motor Circuits and Neuropsychiatry. Boca Raton: CRC Press, pp. 45-88.


Eckenstein, F., and H. Thoenen (1982) Production of specific antisera and monoclonal antibodies to choline acetyl-transferase: characterization and use for identification of cholinergic neurons. EMBO J. 1: 363-368.


Eckenstein, F.P., R.W. Baugham, and J. Quinn (1988) An anatomical study of cholinergic innervation in rat cerebral cortex. Neuroscience 25: 457‑474.


Fallon, J.H., and R.Y. Moore (1978) Catecholamine innervation of the basal forebrain. IV. Topography of the dopamine projection to the basal forebrain and neostriatum. J. Comp. Neurol. 180: 545-580.


Freedman, L.J., and M.D. Cassell (1994) Distribution of dopaminerigc fibers in the central division of the extended amygdala of the rat. Brain Res. 633: 243-252.


Fibiger, H.C. (1991) Cholinergic mechanisms in learning, memory and dementia: a review of recent evidence. Trends Neurosci. 14: 220-223.


Fibiger, H.C., and A.G. Phillips (1986) Reward, motivation, cognition: psychobiology of mesotelencephalic dopamine systems. In: V.B. Mountcastle, F.B. Bloom, and S.R. Geiger (eds.): Handbook of Physiology, Vol. 4, Section 1. Bethesda: American Physiol. Soc. Press, pp. 647-675.


Gaykema, R.P.A., P.G.M. Luiten, C. Nyakas, and J. Traber (1990) Cortical projection patterns of the medial septum‑diagonal band complex. J. Comp. Neurol. 293: 103‑124.


Gerfen, C.R., and P.E. Sawchenko (1984) An anterograde neuroanatomical tracing method that shows the detailed morphology of neurons, their axons and their terminals: immunohistochemical localization of an axonally transported plant lectin, Phaseolus vulgaris leucoagglutinin (PHA‑L). Brain Res. 290: 219‑238.


Gerfen, C.R., M. Herkenham, and J. Thibault (1987) The neostriatal mosaic: II. Patch-and matrix-           directed mesostriatal dopaminergic and non-dopaminergic systems. J. Neurosci. 7: 3915-3934.


Geula, C., and J.T. Slevin (1989) Substantia nigra 6-hydroxydopamine lesions alter dopaminergic synaptic markers in the nucleus basalis magnocellularis and striatum of rats. Synapse 4: 248-253.


Groenewegen, H.J., and H.W. Berendse (1994) Anatomical relationships between the prefrontal cortex and the basal ganglia in the rat. In  A.-M. Thierry et al. (eds): Motor  and Cognitive Functions of the Prefrontal Cortex. Berlin: Springer, pp. 51-77.


Imperato, A., S. Puglisi-Allegra, M.G. Scrocco, P. Casolini, S. Bacchi, and L. Angelucci (1992) Cortical and limbic dopamine and acetylcholine release as neurochemical correlates of emotional arousal in both aversive and non-aversive environmental changes. Neurochem Int. 20: 265S-270S.


Imperato, A., M.C. Obinu, and G.L. Gessa (1993) Stimulation of both dopamine D1 and D2 receptors fascilitate in vivo acetylcholine release in the hippocampus. Brain Res. 618: 341-345.


Ingham, C.A., J.P. Bolam, B.H. Wainer, and A.D. Smith (1985) A correlated light and electron microscopic study of identified cholinergic basal forebrain neurons that project to the cortex in the rat. J. Comp. Neurol. 239: 176‑192.


Itoh, Z., K. Akiva, S. Namura, N. Miguno, Y. Nakamura, and T. Sugimoto (1979) Application of coupled oxidation reaction to electron microscopic demonstration of horseradish peroxidase: cobalt-glucose oxidase method. Brain Res. 175: 341-346.


Jones, B.E., and A.C. Cuello (1989) Afferents to the basal forebrain cholinergic cell area from pontomesencephalic‑catecholamine, serotonin, and acetylcholine‑neurons. Neuroscience 31: 37‑61.


Klitenick, M.A., A.Y. Deutch, L. Churchill, and P.W. Kalivas (1992) Topography and functional role of dopaminergic projections from the ventral mesencephalic tegmentum to the ventral pallidum. Neuroscience 50: 371-386.


Levey, A.I., S.M Hersch, D.B. Rye, R.K. Sunahara, H.B. Niznik, C.A. Kitt, D.L. Price, R. Maggio, M.R. Brann, and B.J. Ciliax (1993) Localization of D1 and D2 dopamine receptors in brain with subtype-specific antibodies. Proc. Natl. Acad. Sci. USA 90: 8861-8865.


Luiten, P.G.M., R.P.A. Gaykema, J. Traber, and D.G. Spencer (1987) Cortical projection patterns of magnocellular basal nucleus subdivisions as revealed by anterogradely transported Phaseolus vulgaris leucoagglutinin. Brain Res. 413: 229‑250.


Martinez-Murillo, R., F. Semenenko, and A.C. Cuello (1988) The origin of tyrosine hydroxylase-immunoreactive fibers in the regions of the nucleus basalis magnocelluaris of the rat. Brain Res. 451: 227-236.


Martinez-Murillo, R., R.M. Villalba, and J. Rodrigo (1990) Immunocytochemical localization of cholinergic terminals in the region of the nucleus basalis magnocellularis of the rat: a correlated light and electron microscopic study. Neuroscience 36: 361-376.


McKinney, M., J.T. Coyle, and L.C. Hedreen (1983) Topographic analysis of the innervation of the rat neocortex and hippocampus by the basal forebrain cholinergic system. J. Comp. Neurol. 217: 103‑121.


Mesulam, M.-M., E.J. Mufson, A.I. Levey, and B.H. Wainer (1983a) Cholinergic innervation of the cortex by the basal forebrain: cytochemistry and cortical connections of the septal area, diagonal band nuclei, nucleus basalis (substantia innominata), and hypothalamus in the rhesus monkey. J. Comp. Neurol. 214: 170‑197.


Mesulam, M.-M., E.J. Mufson, B.H. Wainer, and A.I. Levey (1983b) Central cholinergic pathways in the rat: an overview based on an alternative nomenclature (Ch1‑Ch6). Neuroscience 10: 1185‑1201.


Metherate, R., Ch.L. Cox, and J.H. Ashe (1992) Cellular bases of neocortical activation: modulation of neural oscillations by the nucleus basalis and endogeneous acetylcholine. J. Neurosci. 12: 4701-4711.


Milner, T.S. (1991) Cholinergic neurons in the rat septal complex: ultrastructural characterization and synaptic relations with catecholaminergic terminals. J. Comp. Neurol. 314: 37-54.


Napier, T.C., P. Simson, and B.S. Givens (1991) Dopamine electrophysiology of ventral pallidal/substantia innominata neurons: Comparison with the dorsal globus pallidus. J. Pharm.. Exptl. Ther. 258: 249-262.


Paxinos, G., and C. Watson (1986) The Rat Brain in Stereotaxic Coordinates.  New York: Academic Press.


Perry, E.K., R.H. Perry, G. Blessed, and B.E. Tomlinson (1977) Necropsy evidence of central cholinergic deficits in senile dementia. Lancet 1: 189.


Perry, R.H., E.K. Perry, C.J. Smith, J.H. Xuereb, D. Irving, C.A. Whitford, J.M. Candy, and A.J. Cross (1987) Cortical neuro­pathological and neurochemical substrates of Alzheimer's and Parkinson's diseases. J. Neural Transm. 24 (suppl.): 131-136.

Robertson, G.S., and W.A. Staines (1994) D1 dopamine receptor agonist-induced fos-like immunoreactivity occurs in basal forebrain and mesopontine tegmentum cholinergic neurons and       striatal neurons immunoreactive for neuropeptide Y. Neuroscience 59: 375-387.


Saper, C.B. (1984) Organization of cerebral cortical afferent systems in the rat. II. Magnocellular basal nucleus. J. Comp Neurol. 222: 313‑342.


Schwaber, J.S., W.T. Rogers, K. Satoh, and H.C. Fibiger (1987) Distribution and organization of cholinergic neurons in the rat forebrain demonstrated by computer-aided data acquisi­tion and three dimensional reconstruction. J. Comp. Neurol. 263: 309-325.


Semba, K., P.B. Reiner, E.G. McGeer, and H.C. Fibiger (1988) Brainstem afferents to the magnocellular basal forebrain studied by axonal transport, immunohistochemistry, and electrophysiology in the rat. J. Comp. Neurol. 267: 433‑453.


Smith, Y., B. Lavoie, J. Dumas, and A. Parent (1989) Evidence for a distinct nigropallidal projection in the squirrel monkey. Brain Res. 482: 381-386.


Vertes, R.P. (1988) Brainstem afferents to the basal forebrain in the rat. Neuroscience 24: 907‑935.


Voorn, P., B. Jorritsma-Byham, C. van Dijk, and R.M. Buijs (1986) The dopaminergic innervation of the ventral striatum in the rat: a light- and electron-microscopical study with antibodies against dopamine. J. Comp. Neurol. 251: 84-99.


Wainer, B.H., E.J. Levey, E.J. Mufson, and M.-M. Mesulam (1984) Cholinergic systems in mammalian brain identified with antibodies against choline acetyltransferase. Neurochem. Int. 6: 163‑182.


Whitehouse, P.J., D.L. Price, A.W. Clark, J.T. Coyle, and M.R. DeLong (1981) Alzheimer's disease: evidence for selective loss of cholinergic neurons in the nucleus basalis. Ann. Neurol. 10: 122‑126.


Wood, P.L. (1986) Pharmacological evaluation of GABAergic and glutamatergic inputs to the nucleus basalis‑cortical and the septal‑hippocampal cholinergic projections. Can. J. Physiol. Pharmacol. 64: 325‑328.


Zaborszky, L. (1989) Afferent connections of the forebrain cholinergic projection neurons, with special reference to monoaminergic and peptidergic fibers. In M. Frotscher and U. Misgeld (eds): Central Cholinergic Synaptic Transmission. Basel: Birkhauser, pp. 12-32.


Zaborszky, L. (1992) Synaptic organization of basal forebrain cholinergic projection neurons. In E.D. Levin, M.W. Decker, and L.L. Butcher (eds): Neurotransmitter Interactions and Cognitive Function. Basel: Birkhauser, pp.27-65.


Zaborszly, L., and W.E. Cullinan (1992) Projections from the nucleus accumbens to cholinergic neurons of the ventral pallidum: a correlated light and electron microscopeic double-immunolabeling study in rat. Brain Res. 570: 92-101.


Zaborszky, L. and W.E. Cullinan (1995) Direct catecholaminergic-cholinergic interactions in the basal forebrain. I. Dopamine-b-hydroxylase- and tyrosine  hydroxylase input to cholinergic projection neurons. J. Comp. Neurol.


Zaborszky, L., J. Carlsen, H.R. Brashear, and L. Heimer (1986a) Cholinergic and GABAergic afferents to the olfactory bulb in the rat with special emphasis on the projection neurons in the horizontal limb of the diagonal band. J. Comp. Neurol. 243: 488‑509.


Zaborszky, L., L. Heimer, F. Eckenstein, and C. Leranth (1986b) GABAergic input to cholinergic forebrain neurons: an ultrastructural study using retrograde tracing of HRP and double immunolabeling. J. Comp. Neurol. 250: 282‑295.


Zaborszky, L., W.E. Cullinan, and A.Braun (1991) Afferents to basal forebrain cholinergic projection neurons: an update.  In: T.C. Napier, P.W. Kalivas, and I. Hanin (eds.): The Basal Forebrain: Anatomy to Function. New York: Plenum Press, pp. 43-100.


Zaborszky, L., W.E. Cullinan, and V.N. Luine (1993) Catecholamine-cholinergic interaction in the basal forebrain. Progr. Brain Res. 98: 31-49.




Fig. 1. Computer-assisted reconstruction of PHA-L-filled perikarya in the ventral mesencephalon of 4 representative cases, showing injection sites in progressively more lateral parts of the VTA-SN. Three anterior-posterior levels are shown from left to right for each case. 93128: PHA-L deposit in medial part of the VTA. 93047: Labeled cells are distributed in the lateral half of the VTA and the most medial component of the SNC. See also Fig. 2A. 93049: Injection in the central part of the SN, covering both pars compacta and reticularis. See also  Fig. 2B. 93051: PHA-L-labeled cells in the lateral part of the SNC and SNR.


Fig. 2. Photomicrographs of PHA-L injection sites. A: Tracer-filled cells covering the lateral portion of the VTA and ventromedial corner of the SNC (case 93047). B: PHA-L-labeled neurons in the central part of the SN (case 93049). C-F: Double immunofluorescence labeling of PHA-L (Texas Red, C and E) and TH (Fluorescein, D and F) within the SNC (C,D) and VTA (E,F). A large proportion of the PHA-L labeled cell bodies contained TH within the dopamine cell-rich areas. Scale bar: in A = 50 mm, also applies  for B; C = 50 mm, also applies for D-F.


Fig. 3. Darkfield photomicrographs of PHA-L-labeled fibers in basal forebrain emanating from VTA-SN. Drawings to the right are shown for reference. A: Dense innervation from medial VTA into the HDB. B: Lateral VTA-medial SN afferents innervating the VP and angular part of the VDB. Scale bars in A = 500 µm, B = 200 µm.


Fig. 4. Camera-lucida-assisted line drawings of PHA-L-labeled projections in the basal forebrain ascending from the medial portion of the VTA (case 93128). PHA-L-labeled fibers and boutons (in black) are shown in relation to ChAT-immunoreactive cell bodies (drawn in red) at 4 different posterior-anterior levels (A-D, respectively). Figures were drawn from sections processed for PHA-L-ChAT double-labeling immunocytochemistry.


Fig. 5. Distribution of PHA-L-labeled projections from lateral VTA-medial SNC to basal forebrain relative to cholinergic neurons in sections from case 93047 from posterior (A) to anterior (D). Drawn as in Fig. 4. Asterisks indicate the presence of dense anterograde labeling in the CPu that is omitted in the drawings for simplicity.


Fig. 6. Pattern of PHA-L-labeled (black) and ChAT-positive (red) structures throughout the basal forebrain of case 93049 with PHA-L deposited in central SN.


Fig. 7. Computer-assisted reconstructions of ChAT-immunoreactive neuronal populations of cases 93128 (top row, A, B), 93047 (middle row, C, D), and 93049 (bottom row, E,F). In each image, 12 sections are stacked together. The 3-D images provide a dorsal (A, C, E) and lateral (B, D, F) frontal view of the sections with ChAT-positive cells (white dots) and sites of PHA-L-labeled input (red diamonds).


Fig. 8. Histogram showing the ratio of ChAT-positive cells that may receive PHA-L-labeled input from ventral midbrain to the total number of ChAT-immunoreactive cells of the BFC at each level along the anterior-posterior axis. Diagram shows ratios for each of the four cases (see legend). The x axis shows the A-P coordinates of the sections related to bregma.


Fig. 9. Correlated light and electron microscopy of an axosomatic contact between a PHA-L-labeled bouton emanating from the medial SNC and a cholinergic cell in the VP (case 93048). A: Location of the cell indicated by the star in boxed area, which is shown in B. B: Low-power micrograph of section showing ChAT-positive neurons and plexus of PHA-L-labeled fibers (far left). The recipient neuron is in boxed area. Dark area on the top left is anterior commissure. B': Enlargement of boxed area in B containing the recipient ChAT-positive neuron. C: Low-resolution electron micrograph of the ChAT-positive cell body and PHA-L-filled terminal establishing a synapse. Inset C' shows the terminal abutting the cell (arrow). D: High power of a symmetrical synaptic contact (open arrows) between the afferent fiber (heavily electron dense) and the cholinergic cell. Presynaptic vesicles and mitochondria are electron lucent. Note the flocculent DAB precipitate along the postsynaptic membrane and on the rough endoplasmic reticulum. The PHA-L-positive bouton is neighbored by an unlabeled terminal (asterisk), which shows a small area of asymmetrical synaptic specialization (shown in D', adjacent ultrathin section). Scale bars in B = 200 µm, B' = 50 µm, C = 2 µm, C' = 10 µm, D,D' = 0.5 µm.


Fig. 10. Micrographs of afferent input on a proximal dendrite/cell body of a cholinergic cell in the SI (case 93047). A: Location of the cell indicated by the star in boxed area. B: Low power micrograph of boxed area in A. ChAT-immunoreactive cell in the center (in box) is enlarged in C and D. D: Arrow points to the bouton contacting the dendrite. E: Electron micrograph of the cholinergic neuron. Dendritic shaft, indicated by electron-dense precipitate, extends to the lower right. Bouton is present in boxed area F. F: Detail of the strongly electron-dense PHA-L-labeled terminal. Open arrows indicate the synaptic cleft. Heavy immunoprecipitate at the synaptic membrane obscures possible postsynaptic thickening. Scale bars in B = 200 µm, C = 50 µm, D = 10 µm, E = 2 µm, F = 0.5 µm.


Fig. 11.  Examples of nigral innervation of cholinergic neurons in the basal forebrain. A: Cholinergic neuron in the lateral part of the BST (boxed area) receiving afferent input from lateral VTA/medial SNC (case 93047). B: ChAT-immunoreactive cells in GP-SI area receiving input from the central part of SN (case 93049) with several cholinergic cells in box C. C: Enlarged view of two neurons in B. The dorsal neuron is shown at higher magnification in E, and the ventral neuron is enlarged in F. D: Enlarged view of the neuron from the boxed area in A. Note the color difference between the cholinergic neuron (brown) and the NiDAB labeled bouton (black). E,F: Enlarged view of the boxed areas in C. G: This bouton establishes an asymmetrical synaptic contact (postsynaptic thickening indicated with filled arrows). H, I: Electron micrographs of labeled boutons, shown in E and F, respectively, establishing synapses (open arrows) with ChAT-positive proximal dendritic shaft (H) and cell body (I). Scale bars in A,B,C = 200 mm, D,E,F = 10 mm and G,H,I = 0.5  mm.



3V                   third ventricle

ac                    anterior commissure

acp                  posterior limb of the anterior commissure

Acb                 nucleus accumbens

AcbC              core of the nucleus accumbens

AcbSh shell of the nucleus accumbens

BFC                 basal forebrain cholinergic system

BLA                basolateral nucleus of the amygdala

BST                 bed nucleus of the stria terminalis

CeA                central nucleus of the amygdala

ChAT  choline acetyltransferase

CPu                caudate putamen

DA                  dopamine

DAB                3,3'-diaminobenzidine

f                       fornix

GA                  glutaraldehyde

GABA gamma-amino butyric acid

GP                   globus pallidus

HDB               horizontal limb of the diagonal band

ic                     internal capsule

LH                  lateral hypothalamus

LPO                lateral preoptic area

LS                    lateral septum

LV                   lateral ventricle

MeA               medial nucleus of the amygdala

MS                  medial septum

NiDAB           nickel-enhanced DAB

ot                     optic tract

ox                    optic chiasm

PF                   paraformaldehyde

PHA-L            Phaseolus vulgaris-leucoagglutinin

Pir                   piriform cortex

SI                     substantia innominata

sm                   stria medullaris

SN                   substantia nigra

SNC                substantia nigra pars compacta

SNR                substantia nigra pars reticulata

TH                  tyrosine hydroxylase

Tu                   tuberculum olfactorium

VDB                vertical limb of the diagonal band

VM                 ventromedial nucleus of the thalamus

VP                   ventral pallidum

VTA                ventral tegmental area

ZI                    zona incerta